Upload
trinhdang
View
222
Download
4
Embed Size (px)
Citation preview
Iris recognition: An emerging security environment for human identification
M.Daris Femila #1
, A. Anthony Irudhayaraj #2
1Assistant Professor, Department Of Computer Science, SRM Arts and Science College, Chennai, India
2Dean-Information Technology, Aarupadai Veedu Institute Of Technology, Chennai, India
Abstract
Unlike other biometrics such as fingerprints and face, the
distinct aspect of iris comes from randomly distributed
features. This leads to its high reliability for personal
identification and at the same time, the difficulty in
effectively representing such details in an image. Iris is a
protected internal organ whose random texture is stable
throughout life, it can serve as a kind of living password that
one need not remember but one always carries along.
Because the randomness of iris patterns has very high
dimensionality, recognition decisions are made with
confidence levels high enough to support rapid and reliable
exhaustive searches through national-sized databases. Iris
recognition has shown to be very accurate for human
identification. This paper proposes a technique for iris
pattern extraction utilizing the graph cut method where the
pupilary boundary of the iris is determined. The limbic
boundary is identified by adaptive thresholding method. The
iris normalization was invariant for translation, rotation and
scale after mapping into polar coordinates. The proposed
method has an encouraging performance, success rate of
localization and normalization and reduces the system
operation time. The proposed method involves Graph cut
method, Adaptive thresholding, Normalization modules.
Keywords: iris, pattern, identification, thresholding, pupilary,
normalization
1. Introduction.
Biometrics is the science of measuring physical
properties of living beings. It is a collection of automated
methods to recognize an individual person based upon a
physiological or behavioral characteristic. The characteristics
measured are face, fingerprints, hand geometry, handwriting,
iris, retinal, vein, voice etc. In present technology scenario
biometric technologies are becoming the foundation of an
extensive array of highly secure identification and personal
verification solutions. As the level of security breaches and
transaction fraud increases, the need for highly secure
identification and personal verification technologies is
becoming apparent. Biometrics involves using the different
parts of the body, such as the fingerprint or the eye, as a
password or form of identification. Currently, in crime
Investigations fingerprints from a crime scene are being used
to find a criminal. However, biometrics is becoming more
public. Iris scans are used in United Kingdom at ATM's
instead of the normal codes. In Andhra Pradesh Iris
recognition is being used to issue house hold ration cards.
Practically all biometric systems work in the same
manner. First, a person is enrolled into a database using the
specified method. Information about a certain characteristic of
the human is captured. This information is usually placed
through an algorithm that turns the information into a code
that the database stores. When the person needs to be
identified, the system will take the information about the
person again, translates this new information with the
algorithm, and then compares the new code with the ones in
the database to discover a match and hence, identification.
Biometrics works by unobtrusively matching
patterns of live individuals in real time against enrolled
records. Leading examples are biometric technologies that
recognize and authenticate faces, hands, fingers, signatures,
irises, voices, and fingerprints. Biometric data are separate and
distinct from personal information. Biometric templates
cannot be reverse-engineered to recreate personal information
and they cannot be stolen and used to access personal
information.
2. Iris
The iris has been historically recognized to possess
characteristics unique to each individual. In the mid 1980s,
two ophthalmologists „Dr. Leonard Flom‟ and „Aran Safir‟
proposed the concept that no two irises are alike[6]. They
researched and documented the potential of using the iris for
M Daris Femila et al, Int. J. Comp. Tech. Appl., Vol 2 (6), 3023-3028
IJCTA | NOV-DEC 2011 Available [email protected]
3023
ISSN:2229-6093
identifying people and were awarded a patent in 1987. Soon
after, the intricate and sophisticated algorithm that brought the
concept to reality and it was developed by Dr. John Daugman
and patented in 1994[3].
2.1. Features of the Iris
The human iris is rich in features, can be used to
quantitatively to distinguish one eye from another. The iris
contains many colleagues fibers, contraction furrows, coronas,
crypts, color, serpentine, vasculature, striations, freckles, rifts,
and pits. Measuring the patterns of these features and their
spatial relationships to each other provides other quantifiable
parameters for identification process. The statistical analyses
indicated that the Iridian Technologies IRT process
independent measures of variation to distinguish one iris from
another. It allows iris recognition to identify persons with an
accuracy with a magnitude greater than any other biometric
systems.
2.2.Uniqueness of the Iris
The iris is unique due to the chaotic morphogenesis
of that organ. Dr. John Daugman stated that “An advantage
the iris shares with fingerprints is the chaotic morphogenesis
of its minutiae. The iris texture has chaotic dimension
because its details depend on initial conditions in embryonic
genetic expression; yet, the limitation of partial genetic
penetrance (beyond expression of form, function, color and
general textural quality), ensures that even identical twins
have uncorrelated iris minutiae. Thus the uniqueness of every
iris, including the pair possessed by one individual, parallels
the uniqueness of every fingerprint regardless of whether there
is a common genome”.
2.3.Stability of the recognition
An iris is not normally contaminated with foreign
material, and human instinct being what it is, the iris, or eye,
is one of the most carefully protected organs in one‟s body. In
this environment, and not subject to deleterious effects of
aging, the features of the iris remain stable and fixed from
about one year of age until death. The human eye has
physiological properties that can be exploited to impede use of
images and artificial devices to spoof the system.The iris are
perforated close to its centre by a circular aperture known as
the pupil. The function of the iris is to control the amount of
light entering through the pupil, and this is done by the
sphincter and the dilator muscles, which adjust the size of the
pupil. The average diameter of the iris is 12 mm, and the pupil
size can vary from 10% to 80% of the iris diameter.
The iris consists of a number of layers, the lowest is
the epithelium layer, which contains dense pigmentation cells.
The stromal layer lies above the epithelium layer, and
contains blood vessels, pigment cells and the two iris muscles.
The density of stromal pigmentation determines the colour of
the iris. The externally visible surface of the multilayered iris
contains two zones, which often differ in color. An outer
ciliary zone and an inner pupillary zone, and these two zones
are divided by the collarette – which appears as a zigzag
pattern.
The iris is the plainly visible, colored ring that
surrounds the pupil. It is a muscular structure that controls the
amount of light entering the eye, with intricate details that can
be measured, such as striations, pits, and furrows. The iris is
not to be confused with the retina, which lines the inside of the
back of the eye. Figure1 shows human eye characteristics. No
two irises are alike. There is no detailed correlation between
the iris patterns of even identical twins, or the right and left
eye of an individual. The amount of information that can be
measured in a single iris is much greater than fingerprints, and
the accuracy is greater than DNA.
Iris: This is the colored part of the eye: brown, green, blue,
etc. It is a ring of muscle fibers located behind the cornea and
in front of the lens.
Pupil: Pupil is the hole in the center of the iris that light
passes through. The iris muscles control its size.
Sclera: The sclera is the white, tough wall of the eye. It along
with internal fluid pressure keeps the eyes shape and protects
its delicate internal parts.
Figure 1: Structure of a human eye
Recently,Du et al. designed a local texture analysis
algorithm to calculate the local variances of iris images and
generate a one dimensional iris signature which relaxed the
requirement of entire whole iris for identification and
recognition[7][8].However, all of these algorithms assume
that a circular iris pattern has been successfully extracted from
a captured image but these algorithms are very complex, takes
longer time for code extraction and code matching from the
database. But this paper proposes a new and easy methods for
iris localization and iris normalisation when compared to
other algorithms which are used for iris recognition.
M Daris Femila et al, Int. J. Comp. Tech. Appl., Vol 2 (6), 3023-3028
IJCTA | NOV-DEC 2011 Available [email protected]
3024
ISSN:2229-6093
3. Methodology.
This paper deals with the generation of stable key
from iris image and it is carried over using iris database. The
input image was subjected to segmentation to detect the two
circles i.e., iris/sclera boundary and the iris/pupil boundary.
The resultant image is normalized to produce iris regions. The
proposed method involves three modules namely i)Graph cut
method, ii)Adaptive thresholdingand iii) Normalization.
3.1. Introduction to Iris Recognition
Iris recognition technology combines computer
vision, pattern recognition, statistical inference, and optics. The iris is an externally visible, yet protected organ whose
unique epigenetic pattern remains stable throughout adult life.
These characteristics make it very attractive for use as a
biometric for identifying individuals. Image processing
techniques can be employed to extract the unique iris pattern
from a digitized image of the eye, and encode it into a
biometric template, which can be stored in a database. This
biometric template contains an objective mathematical
representation of the unique information stored in the iris, and
allows comparisons to be made between templates.
When a subject wishes to be identified by an iris
recognition system, their eye is first photographed, and then a
template created for their iris region. This template is then
compared with the other templates stored in a database until
either a matching template is found and the subject is
identified, or no match is found and the subject remains
unidentified. Iris recognition allow user to hands-free
operation in application. Iris recognition has highest proven
accuracy, had no false matches in over two million cross
comparison, according to Biometric Testing Final Report. It
allow high speed also for large populations, just look into a
camera for a few seconds. The iris is stable for each individual
throughout his or her life and do not change with age. The
weaknesses are Intrusive, High cost, Contact lenses ,
sunglasses , optical glasses.
4. Steps involved
The first step towards achieving a homogenous
region is by setting the values of pixels below 60 and above
240 are equal to 255. By doing this we can easily identify the
IRIS boundaries. The purpose for adjusting these values is to
reduce the effect of specularities that may be present in the
pupil. The input is a Captured Eye image and the output is
Homogenized Image.
Figure 2: The overall methodology flowchart
4.1. Iris localization
Iris localization involves Pupilary Boundary
Detection and Limbic Boundary Detection. The methods used
for the detection of PupilaryBoundary(Inner Boundary) is
Graph cut method . Even under the ideal imaging conditions
the pupil boundary is not a perfect circle and in many cases a
small area of the pupil is taken as the iris area by traditional
methods. Although the captured area is small, considering the
fact that most of the iris patterns exist in the collarette area -
which is a small area surrounding the pupil - the error of
inaccurate segmentation will be significant. Therefore a
method to accurately detect the pupil boundary is highly
required. Graph cut method introduced by Y. Boykov [16] is
an efficient segmentation method based on energy
minimization. This method considers the image as a graph and
searches for a cut in the graph that has minimum energy. The
min-cut/max-flow energy minimization method is commonly
used for the purpose of energy minimization [15]. Using the
graph cut based iris segmentation solves the problem of off
angle imaging and also the non circularity of the pupil that is
one of main sources of error in iris recognition known as pupil
error.
The graph cut theory that is used to minimize the
energy function defined to segment the input eye image.
Consider a graph G=(V,E)in which V is the set of nodes and E
is the set of edges that construct the graph. G is called an
undirected graph if the change in the cost function from one
node to another, is direction independent. An example of a
M Daris Femila et al, Int. J. Comp. Tech. Appl., Vol 2 (6), 3023-3028
IJCTA | NOV-DEC 2011 Available [email protected]
3025
ISSN:2229-6093
graph is shown in figure 3. We define two terminals for the
graph – a source (s) and a sink (t). These two terminals are the
main nodes in the graph and are defined by the user. The
maximum cost (weight) in the graph will be given to these
terminal nodes. Nodes other than the terminals are assigned
nonnegative weights being less than or equal to the weights of
the two terminals. Subset C (C ʗ E) is called a cut if it can
divide V into two separate sets S and (where T is equal to V -
S) in a way that s ϵ S and t ϵ T (s and t are the two terminals
of the graph). The cost of a cut is defined as the sum of the
costs of its edges. The minimum cut problem or the problem
of minimizing the cost function is performed by finding the
cut with minimum cost or energy. Cost is defined as
Where ei,j is the edge or link connecting the two vertices i and
j and wi,j is the weight associated with this edge. Several
methods [15] have been introduced to solve the minimum cut
problem in polynomial time. To segment an image using
graph cut method, the pixels of the image are considered as
the nodes of the graph. The edges represent the relationship
between neighboring nodes or pixels and a cut represents a
partitioning of the image constructed via these nodes. Finding
a minimum cut for the image graph results in a partitioning of
the image which is optimal in terms of the defined cost
function for the cut.
Figure 3: An example of a 2D graph showing two terminals named by
source” and “sink”, and the cut separating the regions. Thick lines connect
the terminal to the pixels of the same region, while the thin lines show its
connection to the pixels from the other region (only a few of the links
are shown in the figure)
To segment the image, the terms of a graph such as
the vertices, source, etc. are defined for the image. The pixels
of the image are defined as the vertices of the graph. All
neighboring pairs of pixels of the image are assumed to be
connected to each other with a link and these links are called
the edges. Capacity of each link is defined in terms of the
sharpness of the edge existing between the pixels. The
sharpness of an edge is defined by the difference between
their intensity values. The label O or “object” can be assigned
to a set of pixels to specify the source or object terminal and
the label B or “background” can be assigned to another set
representing the sink or background pixels. The goal is to find
a cut or a set of edges that separates the object and the
background sets in a way that the cut has the minimum cost.
To perform the minimization process the cost or energy
function is defined. The general form of the energy function is
as follows
The Dp cost is defined as
Where, Max is the large positive value that is assigned to sink
and source terminals during the initial labeling process. The
cost function Vp,q is demonstrated as
Where Ip is the intensity value of the pixel p and dist(p,q) is
the distance between pixel p and pixel q. The term σ is the
variance of pixel intensity values inside the object. In the
proposed method a σ value per cluster is calculated for the
whole image and then these values will be used in the rest of
the process.
Figure 4:Original image
The described graph cut segmentation algorithm is
applied to eye images that are taken for iris recognition
purposes to segment the image and detect the pupil boundary
precisely. Knowing the fact that pupil is a dark region in any
eye, one can assume the gray level of its pixels to be close to
zero. Since all regions of the image, except for the eyelashes
and the pupil, have high gray values, there is need for the
effect eyelashes in the picture. The pixels with small gray
level values are marked as potential vertices to be labeled as
M Daris Femila et al, Int. J. Comp. Tech. Appl., Vol 2 (6), 3023-3028
IJCTA | NOV-DEC 2011 Available [email protected]
3026
ISSN:2229-6093
the source or object vertices of the graph. To detect and
eliminate the pixels related to the eyelashes from the pupil
pixels, the method given in [17] is applied. This method uses
the difference between the pixel intensity value and the mean
of the gray level of its neighboring pixels to decide whether it
is an eyelash pixel or not.
By using Adaptive thresholding technique we can
determine the limbic boundary Note that the iris texture is
brighter than the sclera. By finding the difference between
these two regions we find out the limbic boundary value . So
that we can recognize limbic boundary .we get midpoints of a
limbic boundary and radius of the limbic. By calculating the
distance between each pixel coordinates of image and
midpoint coordinates of limbic, comparing with original
radius of limbic we get radius of limbic boundary.
Figure 5:Localized image
4.2.Iris normalization
Iris normalization and enhancement involves
converting the polar coordinate system to Cartesian coordinate
system. Then converting the iris region from Cartesian
coordinates to the normalized non-concentric polar
representation is modeled as I(x(r,ø),y(r, ø))→I(r, ø)
With
x(r, ø) → (1-r)Xp(ø)+rXi(ø)
y(r, ø) →(1-r)Yp(ø)+rYi(ø)
where I(x,y) is the iris region images (x,y) are the orginal
Cartesian coordinates (r, ø) are the corresponding normalized
polar coordinates and Xp,Yp and Xi,Yi are the coordinates of
pupil and iris boundary along o direction .
Note : ø varies from 0 t0 360
r varies from 0 to Ri-Rp where
Ri=Radius of Iris, Rp=radius of pupil
Figure 6:Normalized image
4.3.Pattern Recognition
Texture near the pupilary boundary and limbic
boundary inside the iris has some errors. So we take the
middle row of iris in order to overcome the errors. We convert
the middle row of bits in to hexadecimal code. In our project
the probability of occurring secret code is nearly 1690
. We
consider 360 bits of middle row of an enhanced image.
Convert these bits into hexadecimal code.
F=1001 1110 1100… 0110 1100 0101 1101 1011
9 E C 6 C 5 D B
(360 Bits code)-----(90 Hexadecimal code)
4.4.Pattern Matching
Hexadecimal code is taken from database and
converted into bits. The comparison is done by computing the
HAMMING DISTANCE between the two codes. The
Hamming distance between an Iris code A and another code B
is given by
4.5.Hamming Distance
Given two patterns A and B the sum of disagreeing
bits(sum of the exclusive–OR between) divided by N the total
number of bits in the pattern. If two patterns are derived from
the same iris, the hamming distance between them will be
close to 0.0 then it is accepted or else rejected.
5. Conclusion
Iris boundaries are recognized by using simple
methods and the less complex and faster algorithms than
previous algorithms and it eliminates pupilary noises and
reflections. Homogenization removes specularities of the
pupil., A method based on graph cuts was presented to
segment the pupil region in an eye image for iris recognition
purposes and thus we can recognize pupilary boundary (inner
boundary) accurately.
M Daris Femila et al, Int. J. Comp. Tech. Appl., Vol 2 (6), 3023-3028
IJCTA | NOV-DEC 2011 Available [email protected]
3027
ISSN:2229-6093
Adaptive threshold method can find the limbic radius
and limbic mid-point. By solving these parameters in circle
equation, we can recognize limbic boundary (outer boundary)
accurately. The region between inner and outer boundary is
iris, it is in the polar form and converted into linear form by
converting the polar coordinate system to cartesian coordinate
system, then converting the iris region from Cartesian
coordinates to the normalized nonconcentric polar
representation we get normalized image. By doing
enhancement, the logical image with 360 in length and
breadth is the difference between the outer and inner boundary
is produced. The texture near the limbic and pupilary
boundary inside the iris has some noises due to eyelashes and
eyelids, by taking the middle row of the enhanced image a
secret code is extracted from it. The secret code is converted
into Hexadecimal code of length 90. Hamming code distance
is being used for pattern matching. It can give the 1690
different iris codes . It can overcome the noises caused by
pupil in the image. In this graph cut method only the gray
level information of the images was used to perform the
segmentation. For future work the method can be expanded to
evaluate color images.
REFERENCES
[1] R.C.Gonzalez, Richard E. Woods, Digital Image Processing,
Pearson Education.
[2] J.Forrester , A. Dick , P.McMenamin , and W.lee , The Eye : Basic Scieences in Practice , W B Saunder , London , 2001.
[3] J.Daugman ,” How Iris Rcognitioin Works “ ,IEEE Transaction
on Circuits and System for Video Technology , Vol . 14, N0 . 1,pp . 21-30 ,2004
[4] Y.Du , R. W. Ives , and D . M. Etter , “ Iris recognition “, a
chapter on biometrics the Electrical Engineering Handbook , 3rd Edition , Boca Raton ,FL CRC press , 2004 (in press)
[5] J.D.Woodward , N.M.Orlans ,and P.T.Higgins , Biometrics , the
McGraw- Hill Company , California , U.S.A ,2002.
[6] L.Flom and A.Safir United states patent No .4,641,349 Iris
Recognition System , Washington D.C.:U.S.Government printing
office .
[7] Y.Du , R. W. Ives , D . M. Etter ,T.B.Welch ,and C-Ichang , “One
dimensional approach to Iris recognition “ ,Proceedings of SPIE
Volume 5405 ,pp.237-247,Apr.,2004
[8] Y.Du , R. W. Ives , D . M. Etter ,T.B.Welch “use of One
Dimensional Approach to Iris signature to rank Iris pattern
similarities ,” submitted to Optical Engineering ,2004 .
[9] R.P.Wildes ,J.C.Asmuth , G.l.Green, S.C.Hsu , R.J. Kolczynski ,
J.R.Matey, and S.E.McBride , “A Machine vision system for Iris recognition “,Mach vision application , vol 9,1-8,1996 .
[10] W.W.Boles and B.Boashash ,”A Human Identification
Technique Using Images of the Iris and Wavelet Transform”, Signal Processing, IEEE Transactions on, Vol.46,No.4,1998.
[11] Y.-P. Huang, S.-W. Luo E.-Y. Chen, ”An Efficient Iris
Recognition System”, Proceedings of the First International Conference on Machine Learning and Cybernetics
[12]Ma, Y. Wang, and T. Tan, “Iris Recognition Using Circular
Symmetric Filters”, Pattern Recognition, 16th International Conference on Vol.2,pp.414-417,2002.
[13] Daugman, John G.”Biometric Personal Identification System
Based on Iris Analysis”, U.S.Patent 5,291,560.1994.
[14] Gonzalez, Rafael C., R. E. Woods, and Steven L. Eddins.
“Digital Image Processing using MATLAB.” Prentice Hall,Upper
Saddle River, NJ.20044
[15] Boykov, Y. and Kolmogorov, V., "An Experimental
Comparison of Min-Cut/Max-Flow Algorithms for Energy
Minimization in Vision," in IEEE Trans. on Patt. Anal. and Machine Intel. (PAMI), vol. 26, no.9, pp.1124-1137, Sept. 2004.
[16] Boykov, Y. and Jolly, M.P., "Interactive Graph Cuts for
Optimal Boundary and Region Segmentation of Objects in N-D Images," Proc. Intl. Conf. on Comp. Vision, Vol. I, pp. 105-112,
2001.
[17] H. Mehrabian, A. Poursaberi, B. N. Araabi, “Iris boundary detection for iris recognition using Laplacian mask and Hough
transform”, 4th Machine Vision and Image Processing conference,
Iran, 2007.
[18].A. J. Bron, R. C. Tripathi, and B. J. Tripathi. “Wolff‟s Anatomy
of the Eye and Orbit, 8th Edition.” Chapman and Hall
Medical,London.1997.
M Daris Femila et al, Int. J. Comp. Tech. Appl., Vol 2 (6), 3023-3028
IJCTA | NOV-DEC 2011 Available [email protected]
3028
ISSN:2229-6093